Intelligent Transmitting Surface (ITS)

• Intelligent Transmitting Surface (ITS) is a dynamically programmable two-dimensional array of subwavelength elements that independently tune amplitude and phase of incident electromagnetic waves.
• ITS extends conventional reconfigurable intelligent surfaces by enabling simultaneous transmission and reflection, offering full 360° coverage and improved diversity for next-generation communications.
• Advanced hardware designs and channel models illustrate ITS's potential for enhanced beamforming, energy efficiency, and scalability in complex electromagnetic environments.

An Intelligent Transmitting Surface (ITS) is a dynamically programmable two-dimensional array of subwavelength elements designed to impose independently tunable amplitude and phase transformations on incident electromagnetic waves, predominantly for transmission (refractive) functionality, thereby enabling full-space coverage and advanced channel reconfigurability in electromagnetic environments. ITS technology generalizes the conventional reconfigurable intelligent surface (RIS) paradigm from reflection-only to transmission, and, in its most general form (the STAR-RIS or hybrid mode), supports simultaneous transmission and reflection with per-element-independent control, leading to rich degrees of freedom for beamforming, coverage, and spatial diversity (Xu et al., 2021, Ming et al., 13 Apr 2025, Khalid et al., 2022).

1. ITS Hardware Models and Element Design

An ITS element is typically modeled as a passive or weakly active two-port network characterized by complex transmission ($T_m$) and reflection ($R_m$) coefficients: $$T_m = \sqrt{\beta^T_m}\,e^{j\phi^T_m}, \quad R_m = \sqrt{\beta^R_m}\,e^{j\phi^R_m}$$ where $|T_m|^2 + |R_m|^2 \leq 1$ ensures energy conservation. For pure transmission-mode ITS, $R_m = 0$ and $|T_m| = 1$. Advanced architectures such as the Beyond-Diagonal RIS (BD-RIS) feature independent control of $T_m$ and $R_m$ per cell using hybrid topologies: each cell comprises back-to-back phase-reconfigurable antennas (one for transmission, one for reflection) interconnected via a lossless, varactor-based two-port splitter, enabling power splitting over $-20$ to $+20$~dB in the prototype (Ming et al., 13 Apr 2025).

Implementation options include:

• PIN-diode patch arrays: sub-GHz/low-GHz, coarse phase/amplitude quantization, low cost.
• Varactor/delay-line structures: continuous control, higher insertion loss at large phase shifts.
• Dielectric/graphene metasurfaces: high-frequency operation, fine/continuous control.
• Active amplifying ITS elements: add gain for outdoor-to-indoor or penetration loss scenarios (Xie et al., 2022).

A summary of hardware architectures and their degree of phase/amplitude control:

Architecture	Amplitude/Phase Control	Operational Frequency
PIN-diode patch array	1-2 bits, coupled T/R	<10 GHz
Delay-line/Varactor	Continuous, partly decoupled	1-60 GHz
Graphene metasurface	Continuous, near-independent	100s GHz–THz
BD-RIS hybrid cell	2-bit phase, full power splitting	2-3 GHz (demo), scalable

2. Physical and Analytical Channel Models

ITSs modify the propagation environment by coupling incident waves via subwavelength-distributed, phase- and amplitude-programmable transmission coefficients. Analytical models treat each element as a S-parameterized two-port scatterer, yielding discrete and continuous-space end-to-end channel expressions. Key models:

• Far-field (planewave) approximation: The composite channel between transmitter, surface, and user is $$g_k^T = (\mathbf{r}_k^T)^H\,\mathrm{diag}(T_1,\ldots,T_M)\,\mathbf{h}_\mathrm{Tx \to ITS}$$, modulated by large/small-scale fading and path-loss (Xu et al., 2021, Khalid et al., 2022).
• Near-field/Fresnel diffraction: Applicable for large-aperture ITS or users in the Fresnel region, integrating nonuniform path length and element factor (Xu et al., 2021).
• Huygens–Fresnel/Kirchhoff, polarizability tensor (GSTC), and multiport circuit network: Allow for high-fidelity simulation of beam propagation, near-field effects, and element coupling (Xu et al., 2021).

Path-loss laws, beamforming patterns, and transmission gains must consider the 3D geometry, orientation, and electromagnetic efficiency, including element insertion loss and mutual coupling (Xu et al., 2021, Khalid et al., 2022).

3. Diversity, Beamforming, and Communication Performance

ITS architectures substantially increase communication diversity and beamforming flexibility versus reflect-only RIS:

• Diversity Order: STAR-RIS/ITS structures with $M$ elements achieve a diversity order of $M+1$ per side (for both reflection and transmission), enabling outage and BER slopes unattainable by separate transmit/reflect arrays of the same total aperture. The sum diversity is $2(M+1)$, outstripping the $M+2$ of a conventional transmit/reflect split RIS (Xu et al., 2021).
• Full-space (360°) coverage: By steering transmission and reflection beams independently, ITS achieves full-space beamforming (Ming et al., 13 Apr 2025, Khalid et al., 2022).
• MIMO capacity and precoding: ITS enables hybrid architectures with a small number of active feeds and many passive phase-tunable elements, achieving high spatial multiplexing and array gain with greatly reduced hardware complexity. Precoder design involves non-convex optimization due to unit-modulus and physical constraints, solvable via alternating optimization, SCA, semidefinite relaxation, or manifold-based methods (Khalid et al., 2022, Jamali et al., 2019).
• Zero-forcing/ML-based precoding for ITS: For massive arrays in single-RF-chain scenarios, spatial Sigma–Delta modulation provides low-complexity multiuser precoding—approaching unquantized ZF with only quantized per-element phases (Keung et al., 2023).

Key performance metrics under practical scenarios:

• Beam steering: BD-RIS prototype demonstrates ±45° scan range, HPBW ≈ 8–14°, transmitted/received gain ≈ 6–7 dBi, and up to 50 dB continuous power splitting (Ming et al., 13 Apr 2025).
• Outage/BER: ITS achieves orders-of-magnitude lower outage/BER at a given SNR compared to half-aperture split RIS for the same element count (Xu et al., 2021).
• Energy efficiency: ITS-aided transmitters maintain constant system power as $M$ increases, enabling energy efficiency to rise without additional RF chains (Jamali et al., 2019).

4. Optimization, Control, and Channel Estimation

ITS-based communication requires joint optimization and estimation routines accounting for the unique constraints:

• Element-wise optimization: Joint amplitude and phase control under energy conservation constraints is addressed via alternating optimization (AO), successive convex approximation (SCA), and semidefinite programming (SDP). For coupled transmission/reflection phases in purely passive designs, the per-element phases must satisfy $|\phi^T_m - \phi^R_m| = \pi/2$ or $3\pi/2$ when both $\beta^T_m$ and $\beta^R_m$ are nonzero (Liu et al., 2021).
• Operating protocols: ITSs/STAR-RISs support energy-splitting (ES), mode-switching (MS), and time-switching (TS) modes; resource allocation and switching can be optimized for multi-sector, multi-user, and full-duplex operation (Khalid et al., 2022, Aldababsa et al., 2021).
• Channel estimation: Lack of RF chains at the surface necessitates upstream estimation. Techniques include model-based pilot cycling, compressed sensing (for sparse mmWave), PARAFAC tensor decomposition, deep learning, and reinforcement learning-driven beam training. Efficient two-pilot MAP channel tracking leverages temporal correlation for near-perfect channel state information (CSI) with minimal overhead (Ramezani et al., 27 Jan 2026).
• Scattering matrix and dynamic grouping: In BD-RIS (beyond-diagonal) architectures, dynamic grouping of elements and optimized block-diagonal scattering enhances multiuser sum-rate, implemented by fractional programming and manifold optimization (Li et al., 2022).

5. Prototyping, Practical Realizations, and Limitations

ITS prototypes have validated theoretical predictions in reflection, transmission, and hybrid modes:

• BD-RIS/ITS testbeds: 4×4 BD-RIS at 2.4 GHz achieves dual-mode beamforming, independent transmission/reflection control, switchable in <100 ns via FPGA, with insertion losses <0.8 dB (pure modes), ~4 dB in split mode (Ming et al., 13 Apr 2025).
• Active ITS: Power-amplified transmissive elements, block-grouped for cost reduction, provide high-throughput outdoor-to-indoor penetration at mmWave, with half to an order-of-magnitude fewer elements needed for given sum-rate versus passive designs (Xie et al., 2022).
• Wireless power transfer (WET) ITS: Digital beamforming feeder arrays illuminating passive ITSs enable scalable, energy-efficient RF WET to IoT devices, with SCA-based joint optimization of digital and ITS config (Rosabal et al., 9 Jul 2025).
• Channel estimation for mobility/penetration: ITSs embedded in high-speed railways (transparent windows) are shown to facilitate tractable estimation of Doppler shifts, path gains, and spatial AoDs/AoAs with minimal pilot overhead, outperforming reflection-based solutions in challenging environments (Wang et al., 2022).

Identified practical limitations and non-idealities:

• Phase-amplitude coupling: Passive metasurface elements do not support fully independent amplitude and phase control—actual transfer functions must be characterized empirically per hardware.
• Quantization and mutual coupling: Element discretization and near-field mutual coupling require high-order EM modeling and quantization-aware robust design.
• Insertion loss, complexity, and control: Splitter/varactor and phase shifter insertion loss, finite phase resolution, and large I/O requirement pose implementation challenges, mitigated by block/group amplifying, complexity-reducing protocols, and higher efficiency components.

6. Emerging Applications and Open Research Issues

ITSs are poised for broad impact across upcoming 6G and sub-THz wireless networks:

• Full-duplex and MIMO systems: Simultaneous transmission/reflection with independent beam control enables in-band full-duplex, spatial multiplexed links, and 360° coverage for dense user deployments (Wang et al., 2022, Ming et al., 13 Apr 2025).
• Integrated sensing and communication (ISAC): ITSs support joint radar–communication functions (sensing-assisted comms, vehicle echo localization, dynamic object tracking) using split power and phase control for concurrent echo enhancement and high-rate communication (Meng et al., 2022).
• Wireless power and IoT: ITS-illuminated WET power beacons outcompete conventional analog/digital/hybrid beamforming in scaling, efficiency, and hardware cost for massive IoT deployments (Rosabal et al., 9 Jul 2025).
• Resource allocation and security: Dynamic ATA assignment (grouping), secure programmable transmission, and robust channel estimation for multi-ITS, multi-cell systems remain active research frontiers (Li et al., 2022, Khalid et al., 2022).

Major open challenges include:

• Physically compliant modeling for large surfaces at mmWave/THz bands, encompassing practical nonidealities, delay, and coupling.
• Ultra-low-overhead and real-time channel estimation/beam training, exploiting machine learning while respecting physical and hardware constraints.
• Physical-layer security and reconfigurable jamming in 360° programmable environments.
• Scalable control/management protocols suitable for distributed, federated, and rapidly reconfigurable ITS deployments.
• Active/passive hardware hybridization to approach the theoretical performance bounds of fully adaptive, multi-modal surfaces.

7. Summary Table: Key Properties of ITS/STAR-RIS

Property	ITS/STAR-RIS	Conventional RIS	BD-RIS Prototype
Coverage	360° (dual-sided T & R)	≤180° (reflect-only)	360°, steered in both domains
Per-element control	$|T_m|,\,\angle T_m$; $|R_m|,\,\angle R_m$ (subject to $|T_m|^2+|R_m|^2\le 1$)	$|R_m|,\,\angle R_m$	2-bit phase, continuous power-splitting
Diversity order	$M+1$ per side	$M_\chi+1$ on one side	Up to $2(M+1)$
Hardware complexity	Moderate–High (T/R splitters, per-bit phase)	Low–Moderate	Moderate; demo: 4×4 cells = 160 I/Os
Application domains	Full-duplex, XR/VR, outdoor-to-indoor, WET, ISAC	Coverage extension only	Penetrating glass/wall, FD communication

ITSs, building on reconfigurable surface theory and hybrid transmission/reflection meta-atom design, constitute a foundational enabling platform for the next generation of "smart radio environments," with experimentally validated advantages in spectral efficiency, energy efficiency, and network coverage (Xu et al., 2021, Ming et al., 13 Apr 2025, Khalid et al., 2022).